The evolution of velocity dispersion in the Sco-Cen OB association

Abstract

We study how the stellar velocity dispersion within the Scorpius-Centaurus OB association (Sco-Cen) has evolved over approximately 20 million years, from its formation to the present day, by investigating 32 stellar clusters in Sco-Cen. Using data from the Gaia mission along with supplementary stellar radial velocities, we identified a surprising sequence of abrupt jumps and intervening plateaus in the evolution of velocity dispersion correlating with times of star formation bursts. We find that the association is almost isotropically expanding and that star formation propagated from inside-out with a speed of about 5-6 km/s. We measure a present-day expansion rate of about 10-12 pc/Myr and observe that younger star clusters within the association exhibit higher velocities compared to older ones. This result, along with the stepwise increase in velocity dispersion over time, suggests a structured and sequential star formation process rather than a random one. This phased evolution suggests that stellar feedback is the primary driver of Sco-Cen's star formation history, expansion, and eventual dispersal. Our findings emphasise the value of precisely characterising stellar populations within OB associations, particularly through the creation of detailed, high-resolution age maps.

Figures (10)

• Figure 1: Cumulative 3D velocity dispersion of Sco-Cen, ordered by decreasing cluster age. The orange shaded area shows the 95% interquartile ranges (2$\sigma$ bound), highlighting the uncertainties of the trend (see Appendix \ref{['apx:methods:calc-veldisp']}). The symbols are coloured by the formation time of the youngest cluster included in the cumulative calculation (colour correlates with x-axis). We indicate the Lookback Time at the top of the x-axis, since the cumulative $\sigma_\mathrm{3D}$ could also be interpreted as the evolution of velocity dispersion over time. The four vertical grey bars indicate the four main star formation events in Sco-Cen, marking periods of increased star formation rate, as discussed in Ratzenboeck2023b (see their Fig. 3).
• Figure 2: 3D spatial distribution of clusters in Sco-Cen is shown together with the cumulative $\sigma_\mathrm{3D}$. The seven panels show seven age ranges, which dissect the chronological build-up of Sco-Cen. The upper panels display the cumulative 3D velocity dispersion (as in Fig. \ref{['fig:cumul-vel-disp']}, left panel). Below each graph, seven 3D age maps depict the present-day spatial distribution of clusters (in $XYZ$). The clusters are represented by their enveloping surfaces and colour-coded by age. Each of the seven panels only shows the clusters that formed before the indicated cluster formation times. The shown cluster sizes can be interpreted as upper limits of the region’s size at a given age and not as the true physical extent before the present-day, which was likely smaller (see also Sect. \ref{['sec:discuss-larson']}). The figure illustrates which clusters contribute to which jumps or steps in the cumulative $\sigma_\mathrm{3D}$. Sub-regions and cluster chains are labelled in the final panel. The 3D visualisations originate from the studies Ratzenboeck2023b and MiretRoig2025; interactive versions are available in the respective publications.
• Figure 3: Present-day cumulative size of Sco-Cen, when ordering and adding the stellar cluster members by their decreasing age (without considering orbital trace-backs). The sizes can be seen as upper limits of the region’s size at a given age, but not as the true physical extent before the present-day, which was likely smaller (see Sect. \ref{['sec:discuss-larson']}). The green shaded area shows the 95% interquartile range (2$\sigma$) (see Appendix \ref{['apx:methods:calc-size']}). The colours and grey bars are as in Fig. \ref{['fig:cumul-vel-disp']}.
• Figure 4: Speed--time relation. Relative cluster speed ($v$) versus lookback time ($t$), with the oldest cluster ($e$ Lup) as reference point that is excluded from the linear fit (black, dashed circle). The symbols are colour-coded by formation time (or cluster age, see x-axis) and scaled by number of sources per cluster. The linear fit (grey, solid line) is obtained via bootstrapping, with the median fitting parameters given in the upper left corner. The fit uncertainties are plotted as grey shaded areas (1-2-3-$\sigma$).
• Figure 5: Radial-motion--distance relation. Radial component of the relative cluster motions ($v_r$) versus relative cluster distance ($r$), with $e$ Lup as reference point. The symbols are colour-coded by formation time, as in Fig. \ref{['fig:age-speed-distance']}. A linear fit to the data is shown as a grey solid line, with the best fitting slope given in the upper left corner, and the 1-2-3-$\sigma$ fit uncertainties are plotted as grey shaded areas. The two clusters marked by black, dashed circles ($e$ Lup, $\nu$ Cen) are excluded from the linear fit.